Index of content:
Volume 56, Issue S1, November 1974
- PROGRAM OF THE 88TH MEETING OF THE ACOUSTICAL SOCIETY OF AMERICA
- Session A. Physical Acoustics I: Atmospheric Acoustics
- Invited Papers
56(1974); http://dx.doi.org/10.1121/1.1914055View Description Hide Description
Utilizing kinetic information obtained from acoustical, optical, and opto‐acoustical experiments, a comprehensive description of molecular relaxation in the atmosphere has been developed as the theoretical basis for a method of calculating the absorption of sound in air discussed in the following paper. The effects of atmospheric pressure and sound frequency on absorption evolve easily from the theory. The limitations imposed on theoretical calculations by insufficient energy transfer rate information are discussed. Many of the rates required to predict sound absorption are most easily measured using optical techniques. The similarity of acoustic and high‐energy laser beam propagation through the atmosphere is discussed. [Work supported by the Army Research Office, Durham, North Carolina.]
56(1974); http://dx.doi.org/10.1121/1.1914056View Description Hide Description
The aim is a standard method of calculation with the credibility necessary to support major compromises between economic and environmental benefits desired by society. A two‐relaxation model for the absorption of sound in air is fitted to a maximum of available laboratory and field measurements, as well as the detailed knowledge of basic mechanisms described in the previous paper. In this way, a set of seven fairly simple equations are generated whose form is determined by the physical nature of the mechanisms. These formulas may be used to compute the absorption of a pure tone over the following range of variables to an estimated accuracy of ±(10 + 1.5|t − 20|)%: frequency/pressure 102 to 105 Hz/atm, temperature (t) 0° to +35°C (32°−95°F), relative humidity 10%–100%, pressure 2 atm or less. A method of corrected the result to apply to bands of noise is proposed.
- Contributed Papers
56(1974); http://dx.doi.org/10.1121/1.1914102View Description Hide Description
The attenuation of a 1‐MHz ultrasonic signal has been measured in 72% at temperatures of 300°, 500°, and 700 °K and pressures from 1 atm to 1/100 atm. The measuredattenuation was corrected for classical effects to obtain the attenuation due to rotational relaxation. In addition, the total absorption has been compared to computed values based on a linear combination of viscous,thermal conduction, diffusion, and relaxation to check the constants used for each of these terms. [Work supported by Army Research Office, Durham, North Carolina.]
56(1974); http://dx.doi.org/10.1121/1.1914103View Description Hide Description
The attenuation suffered by a sound wave propagating from an elevated source to the ground, in excess of spherical spreading losses and classical and molecular absorptioneffects, is studied. Reported discrepancies between attenuation measurements made in the field and the predictions of theories based only on absorptioneffects are discussed. It is concluded that atmospheric dynamics(turbulence) is a major contributor to the attenuation of sound waves propagating vertically in the atmosphere. Earlier theories on the attenuation of sound due to scattering have led to predictions of a strong (square‐law) frequency dependence and relatively large magnitudes which are not supported by the available field data. The present theory predicts a square‐law frequency dependence only for the special case of a homogeneous, isotropic medium. For a medium such as the atmosphere, with an inhomogeneous, anisotropic outer scale, a much milder frequency dependence is predicted which approaches a cube‐root dependence for the case of an outer scale of turbulence large compared to half an acoustic wavelength (the general case for audio frequencies in the atmosphere). This result is supported by field data, as are the excess attenuation magnitudes predicted by the present theory.
56(1974); http://dx.doi.org/10.1121/1.1914104View Description Hide Description
Refinement of previous theoretical formulations and numerical computations of pressure waveforms as applied to atmospheric traveling infrasonicwaves could include a description of their asymptotic behavior at high frequencies. In the present paper, calculations based on the W. K. B.J. approximation and similar to those introduced by Haskell [J. Appl. Phys. 22, 157–167 (1951)] are performed to describe the asymptotic behavior of infrasonic guided modes as generated by a nuclear explosion in the atmosphere. The results of these calculations are then matched onto numerical solutions which have been given by Harkrider, Pierce and Posey, and others. It is demonstrated that the use of these asymptotic formulas in conjunction with a computer program which synthesizes infrasonicpressure waveforms has enabled the elimination of problems associated with high‐frequency truncation of numerical integration over frequency. In this way, small spurious high‐frequency oscillations in the computer solutions have been avoided. [Work sponsored by Air Force Cambridge Research Laboratory.]
56(1974); http://dx.doi.org/10.1121/1.1914105View Description Hide Description
Prior theoretical formulations and computational techniques for the prediction of pressure waveforms generated by large explosions in the atmosphere have considered only fully ducted modes. In the present paper, a technique for including weakly leaking guided modes in concert with fully ducted modes is developed. Modification of previous theory includes the extension of the boundary condition at the upper halfspace to include a complex horizontal wavenumber. The major alterations to the computer program Infrasonic Waveforms (as described in report by Pierce and Posey, 1970) incurred consist of the computation of the imaginary part of the newly incorporated complex wavenumber, extension of the normal‐mode dispersion function to lower frequencies, and a second‐order correction factor to the phase velocity. [Work sponsored by Air Force Cambridge Research Laboratories.]
56(1974); http://dx.doi.org/10.1121/1.1914106View Description Hide Description
We set out to model theoretically the spectral features of infrasound observed in the ionosphere and believed to be radiated by severe thunderstorms. We explain the dominant 2–5‐min wave period as the effect of atmospheric filtering; shorter periods are excessively attenuated by absorption in transit to the ionosphere, and longer periods are attenuated in portions of the atmosphere where the waves are evanescent because their frequencies are below the acoustic cutoff. An observed spectral “fine structure” within the 2–5‐min band is explained in terms of resonant interactions between the waves and the atmospheric temperature structure. Accurate quantitative modeling of all these details of the storm‐to‐ionosphere transmission coefficient requires numerical integration of the acoustic‐gravity wave equation, including the effects of ground reflection and partial reflections in atmosphere.
56(1974); http://dx.doi.org/10.1121/1.1914107View Description Hide Description
The phase coherence of an acoustic wave propagating through atmospheric temperature and velocity fluctuations determines the maximum useable range of an atmospheric echosonde. Phase coherencemeasurements were made for sound propagating through atmospheric turbulence from a source mounted on top of a 500‐ft meterological tower. The measurements were made in conjunction with other atmospheric acoustic and gravity wave experiments, including operation of a bistatic echosonde. Throughout the observations, instrumentation on the tower at various heights provided mean winds and temperatures and values of the turbulence structure constants CT and Cv . The results are compared with theoretical predictions and with the few previous experimental observations.
56(1974); http://dx.doi.org/10.1121/1.1914108View Description Hide Description
A theoretical analysis of the propagation of spherical time‐harmonic waves in a random medium is presented. The smoothing method is used to study the incoherent (or randomly fluctuating) component of the wave field. With the aid of the Fresnel approximation, an expression for the second moment of the incoherent wave is obtained which is valid for the case of high‐frequency waves propagating in a medium which is statistically homogeneous and isotropic. This expression shows that the rms fluctuations of the wave field increase initially as the square root of the propagation distance, but that at larger distances (and higher frequencies) the fluctuations tend to saturate. These results agree with observations of waves propagating in real media.
- Session B. Physiological Acoustics I: Head Diffraction, Middle Ear and Bone Conduction
56(1974); http://dx.doi.org/10.1121/1.1914162View Description Hide Description
The wave properties of ten ears under blocked‐meatus conditions have been determined using the progressive wave source described earlier (Denver, 1971). Physical models of the ear with four or more modes matching the average ear can be readily constructed provided that the crus helias and fossa are properly represented. These and other anatomical structures of the human ear are clearly associated with specific acoustic parameters. For example, the presence of the crus helias substantially reduces the frequency of the second mode (∼7.1 kHz) without affecting the fourth mode (∼ 12.1 kHz), which is largely dependent on the breadth of the concha. The third mode at 9.6 kHz, overlooked in earlier work, owes its existence to the fossa. Differences in coupling between the fossa and the cymba account for some of the intersubject differences observed at 10 kHz. Physical models of the complete human ear, useable up to 13 kHz, now appear feasible.
56(1974); http://dx.doi.org/10.1121/1.1914163View Description Hide Description
The pressure variation in the region of the ear near the surface of the head was measured on KEMAR, a manikin especially constructed to duplicate the acoustic performance of the human head and body. The purpose of the experiment was to compare the response in various locations so that a single curve could be selected as being representative and used as the free field to human‐head correction curve. Twenty‐five curves on a 5 × 5 matrix with a 2‐cm separation were recorded. Results indicated a variation with both frequency and spatial location. An attempt was made, therefore, to mathematically represent the data in a single equation using frequency and spatial location as variables This would serve as both a means for interpolation between curves and as a mathematical model for the acoustic interactions involved.
56(1974); http://dx.doi.org/10.1121/1.1914164View Description Hide Description
The eardrum has been modelled as a plane, isotropic membrane. We have included the effects of the embedded malleus, the frequency‐dependent reactive and resistive forces exerted on the malleus by the rest of the ossicles, and the closed middle‐ear air cavity. The variations in the form of the drum and malleus among different species have also been considered. The simulation is done using an adaptation of the finite‐element method, in which the two‐dimensional continuum is divided into a number of arbitrarily shaped triangles. The results obtained are compared to recent servations of eardrum vibration in the cat and guinea pig, as well as to middle‐ear impedance data. [Supported by the Medical Research Council of Canada, the Quebec Department of Education, and the McConnell Foundation.]
56(1974); http://dx.doi.org/10.1121/1.1914165View Description Hide Description
A small perforation (diam 1 mm) was placed in the postero‐superior quadrant of the feline tympanic membrane (TM). Together with the voltages across the transducer,sound pressures (amplitudes and phases) required for a 10‐μV CM response (round window) were measured over a range from 200 to 4000 Hz (a) separately in front and behind the TM, (b) in open and closed sound systems, and (c) before and after perforation. Sound‐pressure changes in front of the TM after perforation revealed low‐frequency losses, identical in shape and magnitude for open and closed systems: 12 dB/octave slopes below 1600 Hz. The voltage changes paralleled the sound‐pressure changes only in the open system. Considerably larger voltages were required in the closed system, especially between 630 and 2000 Hz (maximally 18 dB at 1250 Hz). Furthermore, from the sound pressuresmeasured on both sides of the TM, the net force acting upon the TM was calculated. The change in net force after perforation followed an approximate slope of only 6 dB/octave. Implications of these findings will be discussed. [Supported by NIH and HEW grants.]
56(1974); http://dx.doi.org/10.1121/1.1914166View Description Hide Description
The tympano‐ossicular system has been anatomically analyzed in 26 species of heteromyid rodents. Based on anatomical measurements, 24 of the 26 species should have a transmission efficiency of 94%–100% from tympanic membrane to cochlear, at the resonant frequency. The areal ratio of stapes footplate to 2/3 tympanic membrane is remarkably constant among the species, varying only from 0.04 to 0.07: in Dipodomys and Microdipodops this small ratio is due to the very large tympanic membrane; in Perognathus and Liomys it is due to the extremely small stapes footplate. The lever ratio of incus to malleus varies from 0.28 to 0.33 in Dipodomys and Microdipodops, from 0.37 to 0.46 in Perognathus, and from 0.55 to 0.60 in Liomys. In addition, the middle‐ear volumes and the morphology of tympanic membrane, ossicles, ligaments, and muscles all combine to minimize both mass and stiffness. All these data suggest middle‐ear mechanisms which are very efficient over a broad frequency range. We have cochlear microphonic data consistent with this conclusion. [Supported by NIH Grant 11459.]
56(1974); http://dx.doi.org/10.1121/1.1914210View Description Hide Description
The middle‐ear transfer function of the Mongolian gerbil (Meriones unguiculatus) was obtained using a constant CM criterion with both single electrodes in scala media and differential electrodes in scalae media and tympani. A cancellation technique allowed accurate phase measurement of both CM and ear‐canal sound pressure. Results indicate small inter‐animal variability and no significant differences between single electrodes and differential electrodes, except for contamination of single electrode recordings by whole‐nerve action potentials. With the bulla open, the magnitude of the transfer function has a positive slope of 6 dB/octave up to 200 Hz and is flat from 200 Hz to 1.8, where it begins to fall off at a rate of approximately 12 dB/octave. The CM phase leads to that of the sound pressure by 90° at 100 Hz, is 0° at 600 Hz, and lags by 90° at 1.8 kHz. At high frequencies, the phase lag is greater than 180°. The high‐frequency rolloff, steeper than that found in other mammals, is probably due to decreased coupling between the incus and the stapes.
56(1974); http://dx.doi.org/10.1121/1.1914211View Description Hide Description
Properties of acoustic middle‐ear‐muscle (MEM) reflex in Myotis lucifugus were electrophysiologically studied. (1) The shortest latency of the reflex in terms of EMG was 3.4 msec for the stapedius muscle (SM) and 4.4 msec for the tensor tympani muscle. (2) Unlike first‐ and second‐order auditory neurons, all tuning curves of single SM fibers were very broad. Motoneurons for the muscle appeared to integrate signals from many superior olivary neurons with different best frequencies to produce such broad tuning curves. (3) The threshold of the reflex in terms of EMG decreased with an increase in stimulus duration. With a long tone burst, the lowest threshold obtained was 20 dB SPL in both muscles. (4) The MEMs contracted maximally at the beginning of the reflex and then tonically contracted to some extent during the stimulus. The feedback gain of the reflex in terms of cochlear microphonic was 1.0 at the peak of the MEM contraction and 0.3–0.5 during a quasisteady contraction. (5) A frequency‐attenuation curve of the reflex showed that attenuation was maximal at about 30 kHz and reduced with a slope of 17 dB/octave at above it. [Supported by NSF grant GB‐40018.]
56(1974); http://dx.doi.org/10.1121/1.1914212View Description Hide Description
Bone‐conduction (BC) thresholds measured in terms of acceleration had shown large systematic variations, mainly with contact area [J. Acoust. Soc. Am. 55, S62 (1974)]. Sound pressuresgenerated in the occluded ear canal by BC signals of threshold strength were now found to be independent of contact area and, below 2 kHz, equal to those required for AC thresholds. Air/bone cancellation experiments revealed that, below 2 kHz and with the ear canal occluded, the BC stimulus is really given by the sound pressuregenerated in the ear canal, i.e., one measures the response to an air signal instead of that to a bone signal. [Essentially the same conclusion was reached for the cat; cf. Acta Otolaryng. Suppl. 132 (1966).] The cancellation experiments permitted the assessment of BC components unaffected by the sound pressure in the ear canal. These true inner‐ear BC responses also varied with the contact area, demonstrating once more that acceleration is a poor indicator of BC responses. Experiments now in progress will determine if the above findings apply also to the unoccluded ear canal. [Supported by NIH and DRF grants.]
56(1974); http://dx.doi.org/10.1121/1.1914214View Description Hide Description
Acoustic energy delivered to one ear can, through cross conduction, stimulate the contralateral cochlea. The effects of cross conduction can complicate studies attempting to define binaural mechanisms. For instance, acoustic energy delivered directly to one ear can interact with the cross‐conducted acoustic energy delivered to the other ear, thus confounding the independence of the two auditory channels. This constraint upon the independence of the acoustic channels might be more severe for small animals such as the guinea pig and chinchilla that for larger animals such as the cat, monkey, and man. We have determined isopotential curves (Turn I, differential electrodes) for the guinea pig and chinchilla from which interaural attenuation has been estimated. Below 10 kHz, interaural attenuation decreases from about 50–60 dB below 1 kHz to about 40–50 dB around 10 kHz. Differences in curves due to open and closed bullae appear for some animals. For ipsilateral stimulus levels greater than 90 dB SPL, frequencies above 10 kHz produced direct electrical pickup in differential electrodes whether they were into the cochlea or suspended in a saline‐filled bulla. Contralateral stimulus levels above 100 dB produced similar phenomena.
- Session C. Architectural Acoustics IA and IB
56(1974); http://dx.doi.org/10.1121/1.1914262View Description Hide Description
Knowledge of the room constant of an interior space is usually required whenever engineering noise controlmeasures are to be contemplated. Easily usable and transportable experimental methods to determine the room constant are therefore desirable. One convenient method is the balloon impulse method where the impulse produced by bursting a toy balloon is used to produce a high‐level broad‐band excitation for decay time measurements. Extensive use of the balloon impulse method has been made in determining the room constants of beth small (typical office spaces of 5000 to 10 000 ft3) and large (200 000‐ft3 factory spaces and larger) spaces. The balloon impulse method has produced results consistent with values obtained from the steady‐state large‐room‐volume equation using steady‐state decay field sound‐pressure levels measured at specified distances. However, the lack of diffuse field conditions and the nature of the true directivity factor had to be considered in the comparison. Room constant determinations made solely from calculations based on estimated sabin absorption coefficients were found risky. Serious errors are possible if misjudgment of surface properties is made. This is particularly true in large volume spaces.
56(1974); http://dx.doi.org/10.1121/1.1914263View Description Hide Description
Spatial mean‐square sound pressure due to steady sine‐wave excitation was measured in a room whose volume is 65 m3 (2300 ft3) and reverberation time about 0.6 sec. A multi‐direction sound field was measured by rotating a random‐incidence electretmicrophone one revolution in one minute on a circle of 1‐m radius, at the center of the room, mounted either on a boom or on the experimenter, who walked around the circle just back of the boom. Measurements of mean‐square sound‐pressure level were quite consistent (often within 0.5 dB) despite relative interference minima like −40 dB and movement of the experimenter during a rotation. A noise‐average meter made by Computer Engineering Ltd. was used, modified to measure the level of the time integral of squared sound pressure called exposure level. When the microphone was at the breast pocket of a thin shirt worn by the experimenter walking around the circle, the mean‐square sound‐pressure level was greater than it was on the same circle in the absence of the experimenter, in round numbers, by 1, 2, 3, 2, 0, and −1 dB, respectively, at 125, 250, 500, 1000, 2000, and 4000 Hz. The result was nearly the same for a heavy wool jacket and hat, except −4 dB at 4000 Hz. On a shoulder or at an ear, the buildup at the six frequencies was quite similar.